Everyone's still debating this AI architecture question

LangChain · Advanced ·🤖 AI Agents & Automation ·2h ago
Agent inside the sandbox or outside it? Ramp's Head of Applied Research explains why they put the agent outside, spin up a fresh sandbox for every user action, and why keeping code-gen isolated from prior runs was the key architectural insight behind Ramp Sheets. This clip is from Max Agency, a #podcast about how the best AI agents are actually being built. Hosted by Harrison Chase, CEO of LangChain, each episode goes deep with the builders designing, deploying, and learning from real agent systems in the wild. From architecture decisions to evals, tooling, and failure modes, Max Agency is for people who want to understand what it really takes to build useful agents.
Watch on YouTube ↗ (saves to browser)
Sign in to unlock AI tutor explanation · ⚡30

Related AI Lessons

Agent cost bugs are debugging bugs
Agent cost bugs can be caused by unexplainable runs, not just high bills, and require a different approach to debugging
Dev.to AI
Agent Boosting: The Missing Workflow for Getting Real Results from AI Coding Agents
Learn how to bridge the gap between AI coding agents' theoretical capabilities and practical results with Agent Boosting
Dev.to AI
Building an Open-Source Consensus Protocol for Multi-Agent AI — Architecture Decisions and Trade-offs
Learn how to build an open-source consensus protocol for multi-agent AI systems, including key architecture decisions and trade-offs
Dev.to · Shyam Desigan
The 502 Bad Gateway Nightmare: Saving Your AI Applications with Asynchronous Queues
Learn to save your AI applications from the 502 Bad Gateway nightmare by using asynchronous queues, a crucial step in ensuring the reliability and scalability of Large Language Models
Medium · DevOps
Up next
Paperclip + Hermes Agent Is INSANE!
Julian Goldie SEO
Watch →